Energy balance

Associations between specific nutrient intakes and disease are likely secondary to differences in total caloric intake, body size, physical activity, or metabolic efficiency, rather than representing primary etiologic effects of the nutrient itself.

When evaluating diet-disease links, do not trust absolute nutrient numbers (like grams of fat) or simple percentages (nutrient density) if total calories vary between groups. Instead, use statistical methods (like regression residuals) to isolate the nutrient's effect independent of total energy intake. If a nutrient's association with disease disappears after adjusting for calories, the nutrient itself is likely not the primary cause.